AI Will Revolutionise How You Travel, Priceline CEO Says

With its 1997 launch as a name-your-own-price site, Priceline helped usher in the era of online travel booking―and the growth of a trillion-dollar category. Priceline itself has mushroomed into a global business with 1,500 full-time employees; along with global airline bookings, the site claims to offer 1.2 million accommodations in 116 countries.

Now, the Connecticut-based company is focusing on generative AI, and CEO Brett Keller is behind the leap. It’s just the latest technology push from Keller, a 25-year veteran of the company who has also served as CMO and COO. Under Keller, Priceline launched the travel industry’s first full-service mobile app in 2009; he also helped conceive the hugely popular William Shatner-fronted “Negotiator” ad campaign.

Keller, 56, talked to Penta about how AI is changing travel planning, what luxury travellers do to save money, and why Japan blows his mind.

Penta: Much of Priceline’s marketing is about value. Do you see luxury travellers in your customer mix along with budget-conscious travellers?

Brett Keller: High-net-worth consumers take a significantly higher number of trips than the average leisure traveller. For a high-end vacation like a safari in Tanzania, they’ll work with experts. But for the other 30 trips they book that year, either for themselves or family members, they don’t always reach for the stars. They want to manage money effectively. And as they’re moving around the country and the world, they need a fast, easy way to book travel that accommodates their needs. Priceline is a great platform for that kind of trip. And if you’re taking a quick weekend trip to Miami, and want to stay in the Four Seasons, we’ve negotiated with them. Even high-net-worth individuals seek value.

Priceline made headlines last year for partnering with both Google and OpenAI on Penny, your AI assistant. How has Penny evolved?

She’s gone from, “I can answer questions about the hotel you’re looking at” to actually servicing the customer through more complicated scenarios. If you need to cancel a hotel that’s fully refundable, you can do it with one click. But if there are issues, like a reservation that’s not fully cancelable, Penny can walk through those steps for you.

Penny’s also helping people find, search, and book properties. For example, if the customer tells her, “I’m looking for a great resort with these features, anywhere warm”—she’ll present recommended properties that meet those requirements. You can continue with Penny on the site, or go with the traditional experience.

Has there been any pushback from consumers about Penny and generative AI?

There has been none. Customers still have access to phone agents. Anyway, the younger generation doesn’t want to talk to anyone. And with traditional chat agents, live agents, or even messaging apps, there’s typically a delay. Penny answers immediately and in real time.

What’s the future of generative AI and online travel booking?

The future is a highly personalised shopping experience. It’s hard to achieve, because we don’t know about the consumer when they come in. But generative AI lets us dramatically improve personalisation. As you work with Penny, and tell us your preferences, the way you interact lets us find and book the best products and services every time you return. The ability to customise and personalise increases exponentially.

Is the travel experience even more bifurcated between elite, ultra-high-net-worth travellers and everyone else?

Consumers seem to be a little less sensitive in some areas. On planes, first-class and comfort-plus seats are the first to go, most of the time, and people are burning through points to sit in the front because they’re tired of not having legroom. But seats are packed in economy, too, so people are flying.

Hotel bookings are more economically driven in the U.S. Higher-income people are not as affected by interest rates or the cost of living, so they spend more freely than economy-minded consumers. There is a bit of bifurcation there. The low end is not filling, but the high end is.

It’s been reported Europe’s going to get even more crowded this summer. Does Priceline ever suggest alternatives to over-touristed or overpriced destinations?

We’re not in the business of telling you where you shouldn’t travel. We market popular destinations because that’s where people want to go. As much as we could tell people, “Las Vegas is overcrowded, don’t go,” people will want to go.

Social media is highlighting some overpopulated destinations and suggesting alternatives, so that comes back to us. But price is the No. 1 motivator. Vegas is a great value. There are so many hotel rooms available that you can go in a non-peak period and get a room in a four-star hotel for US$120. Try doing that in New York City.

What destinations are going to pop over the next year or two?

Asia will continue to be exciting and interesting to people. Bangkok is a great place to move in, then travel throughout Southeast Asia. As the region gets more popular, people will keep trying to find more remote and more unique destinations. People love Europe, but it was the hot spot in 2022 and ’23. Some travellers are saying, “I’ve had enough, and I’m moving on.” Japan is also amazing, for so many reasons. It’s easy to navigate. English is not a challenge. It’s safe. It feels like a different world, but completely first-world. The strength of the dollar is also driving some of that―again, price plays a role.

Beyond that, Mexico and the Caribbean took a real hit in 2022 and ’23 after a boom in 2021. They’re coming back now. A lot of people don’t want to travel far—“I just want to go to a beach and not think of anything.” And because not everyone wants to travel overseas, unique and relaxing cities like Nashville, Tennessee; Houston; and Austin, Texas, will be popular, especially into the fall.

Every day brings more headlines about airline woes. Who gets blamed if a Priceline customer has a bad flying experience, you or the airline?

When you have a bad experience traveling, you want to blame everybody. No matter what happened, the online travel agency takes blame and the airline takes the blame. It could be your seat, the person sitting next to you, whatever. We get the complaint, and we take on that responsibility and that role. We have leverage because of the amount of business we drive to partners. They want to work with us to make sure the customer has the best experience. Something goes wrong almost every time you take a trip. That’s just the reality.

What are your favourite places to travel?

My favourite destination, and a place where I spend a lot of time every summer, is [resort town] McCall, Idaho, one of the most beautiful towns in the West, with hiking, trails, and mountain biking. Outside of the country, it’s Japan, absolutely. Tokyo is the most exciting city in the world. It’s mind-blowing.

This interview has been edited for length and clarity.

OpenAI says it is scrapping the release of its next-generation AI model over safety concerns that researchers raised during internal testing, in one of the clearest signs so far that agent misbehavior could stymie the industry’s rapid progression.

The move follows a summer punctuated by reports of artificial-intelligence systems industrywide going rogue, and marks a rare case of a major AI developer ditching a new release because of safety concerns.

The company had planned to launch the model, known as GPT-6.1 Astra, in the coming days or weeks, aiming for an October debut. The model was more capable than the company’s previous models in completing challenging tasks from end-to-end without human assistance, as well as writing.

The company instead will focus on improving the safety of future models, which it expects to be even more capable.

Saachi Jain, OpenAI’s head of safety systems, said in an interview that GPT-6.1 Astra regressed in two areas. Compared with its predecessor, GPT-6 Astra, the model performed poorly on tests measuring alignment, or how well the model adheres to what humans would like it to do. Specifically, GPT-6.1 Astra showed higher levels of deception: It wasn’t always honest about telling users of the actions it did or didn’t take.

Another issue was what OpenAI calls “scope authorization,” meaning that GPT-6.1 Astra would push ahead on a task without asking the user for permission, and would at times reach for external tools and services even if it might be unsafe.

“For anything regarding safety and alignment, there’s a trade off,” Jain said. “You really do need to find what’s the right line between staying within scope, but also avoiding laziness in terms of how the model actually pursues tasks even when it hits friction.”

OpenAI CEO Sam Altman attended a United Nations Security Council meeting about AI last week. Alexi J. Rosenfeld/Getty Images

While GPT-6.1 Astra improved in areas such as “model laziness,” Jain said it didn’t quite meet OpenAI’s bar for safety and alignment, so the company decided not to launch the model publicly.

The announcement comes one day ahead of OpenAI’s annual developer conference in San Francisco. In the past, OpenAI has used the conference as an opportunity to launch new models and services that reduce costs for software developers—a segment the ChatGPT-maker competes with rival AI company Anthropic to win over.

In recent weeks, OpenAI and Anthropic have called on industry partners to slow down the development of cutting-edge AI models and invest in safety standards, noting they will temper the pace of their own internal AI progress.

OpenAI says it is working to investigate a range of agent security incidents that it has discovered in recent months, and address the safety issues underneath them. As part of the work, the company has implemented a new monitoring system to catch AI-agent misbehavior more quickly, and started requiring engineers to use stronger security guardrails for testing its AI systems.

Earlier this summer hundreds of OpenAI’s internal agents, which were tasked with completing a cybersecurity test, ended up hacking into the AI company Hugging Face. Since then, high-profile organizations such as the Australian government and United Nations discovered that OpenAI’s agents used similar, but less extensive, techniques to gain access to their websites.

Many of the publicly known agent-security incidents involved OpenAI’s internal AI models that were never slated for public release.

Last week, OpenAI said it paused training on its most capable AI models after an AI agent slipped through a gap in the company’s internet restrictions to query a public chatbot. The company said its new monitoring systems flagged the incident within 15 minutes, and training on these models remains paused.

GPT-6.1 Astra isn’t one of those models, but a different case, the company said.

“We want to make sure our model development is safe no matter whether that’s in the company, or when we ship it to users,” Jain said. “But when we ship it to users, we have an extremely high bar in terms of safety and alignment.”

While the company decided not to ship GPT-6.1 Astra, it hopes to use the same base model to do additional reinforcement learning runs, and create future generations of its GPT-6 models.

OpenAI plans to conduct several deep dives to identify the root cause of the problems identified in GPT-6.1 Astra, Jain said. The work includes ensuring that OpenAI’s reinforcement learning environments are rewarding the right type of behavior, Jain added, though she noted the company would investigate all stages of model development.

AI companies have begun to draw scrutiny from policymakers and public officials, who are paying attention to the rapid development of the technology. Later this week, a Senate subcommittee is holding a hearing with third party AI researchers titled, “Rogue AI: Securing the Homeland Against AI Agent Attacks.”

Florida Attorney General James Uthmeier, a Republican, sued OpenAI in June, claiming that the company and Chief Executive Sam Altman knowingly released an unsafe product and ignored warnings that it could harm users.

In a motion for temporary injunction filed Monday, Uthmeier sought to prevent OpenAI from developing new AI models without third-party approved safeguards, stop ChatGPT from soliciting user engagement and limit the company’s ability to advertise ChatGPT as safe.

Tech companies claim they “cannot stop barreling forward with their potentially civilization-ending endeavors unless they are forced to do so by the government,” Uthmeier said in the filing. “The Florida Attorney General is answering your cry for help.”

An OpenAI spokeswoman said that people want to know AI is being developed safely, “and that starts with what companies like ours do ourselves.”

“Governments have an important role to play in setting robust safety standards for AI, and we’re committed to working with Florida and other states on advancing pragmatic AI policies that apply to the entire AI industry—not just one company,” she said.